Download Review8_103 - Montana State University

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Overexploitation wikipedia , lookup

Latitudinal gradients in species diversity wikipedia , lookup

Sustainable forest management wikipedia , lookup

Habitat wikipedia , lookup

Habitat conservation wikipedia , lookup

Old-growth forest wikipedia , lookup

Tropical rainforest wikipedia , lookup

Forest wikipedia , lookup

Conservation movement wikipedia , lookup

Reforestation wikipedia , lookup

Biological Dynamics of Forest Fragments Project wikipedia , lookup

Transcript
08-103
Assigning Editor Comments:
My apologies to the authors for taking so long to get this back to them. This revised
version of the manuscript was sent to 2 new reviewers as well as back to one of the
original reviewers. The third review, for which we waited a considerable time, was
cursory. The other two reviews raised concerns regarding the manuscript and
unfortunately these concerns seem consistent in many respects to those raised in the first
version.
Reviewer # 1:
General Comments:
Normally I would start my review by pointing out the merits of the paper. Except for the
writing style which is very good, this paper has the quality of what one might expect
from a first draft of a masters thesis proposal. The arguments are poorly thought out and
inconsistent. The justification for the work is tenuous. The data are inadequate and the
analysis correspondingly weak. Results are mostly inconclusive when one only includes
studies with sufficient sampling in the analysis. The authors did not take into account
important differences in context such as differences in the type of edge (e.g. clear cut
versus agriculture). Analysis and interpretation of the data fail to discuss alternative
explanations.
The discussion only mentions the authors’ favorite causal explanation for a correlation
that could be caused by numerous mechanisms. Conservation implications of any import
or novelty are lacking. Basic facts like which biome a study location was in were
incorrect in the tables. Savannas should not be confused with closed canopy forests when
investigating edge effects. The ms has two introductions, one of them presumably a draft,
which implies that none of the authors had the courtesy to proof read the ms before
submission. Aside from a submission to the Caribbean Journal of Science written by a
non-English speaker, this is the worst paper that I have ever reviewed. A senior professor
should know better than to submit such poor quality science to the premier conservation
journal.
Page 10: Using an evolutionary explanation assumes that such edges are naturally
present. If the edges are a recent man-made phenomenon in a region, one would not
expect edge-specialist species. Perhaps a better hypothesis is that ecosystems with little
history of edges and open habitats should have more species intolerant of the stresses
associated with open habitats. It is not a matter of steepness of gradients but consequence
of lacking adaptation because of no previous exposure.
Page 11: Forest fragmentation is not just a threat because it reduces habitat for forest
interior species. It also reduces connectivity and associated dispersal and gene flow.
Smaller patches resulting from fragmentation may not support minimum viable
populations. Effects of climate change and species invasions are also greater in
fragmented areas. There is also a big difference between temporary edges around boreal
forest clear cuts and permanent fragmentation associated with development in temperate
regions or longterm edges associated with forest conversion to agriculture in the tropics.
Consequently managers will continue to be concerned about fragmentation even in areas
with few forest interior specialists. Conservation and management in practice cannot be
separated from important nuances in context if it is to be successful.
Page 14: What about agriculture versus clear cuts? Clear cuts are good wildlife habitat,
agriculture is not. These two types of edges should be analyzed separately. For instance,
agriculture has high levels of human activity such as burning, with fires spreading into
forest edges. Boreal forest clear cuts are rarely visited by humans and contain native
forest vegetation.
Page 16, line 3: I disagree. The tropical studies only sampled a small fraction of local
diversity in many cases and it is likely that the species chosen were not a random subsample. It is certainly not possible to conclude from a study with 4 species about the
effects on a whole community of dozens or even hundreds of species (in the case of
beetles). It is likely that investigators chose species that are likely to show an edge effect.
It appears that his bias is greater in the tropics where communities were seriously undersampled. The analysis should be repeated using only the studies that have a decent
number of species sampled (a minimum of 8 or even 10).
Page 18: That microclimate is different inside the forest is well established among forest
meteorologists. I think that citing the forest meteorology literature could make these
conclusions less tentative. At present only forest ecology papers are cited but appropriate
forest meteorology papers should also be cited.
Page 18: This is a good example where confounding high latitude clear cut edges with
low latitude agricultural field edges is a problem. Boreal forest mammals like moose find
excellent forage in clear cuts, while moose would avoid a cow pasture or soybean field.
Similarly not all mammals are alike. The boreal forest has mostly ground dwelling
mammals like moose and beaver, whereas tropical forests have many arboreal mammals,
which obviously depend on trees. Thus there is a latitudinal difference in species guilds
that affects edge effects independent of biomass. Tree dwelling marsupials of the tropical
rainforest in Queensland Australia will not enter a cow pasture, regardless of the tree
biomass in the forest. Such latitudinal differences in species guild composition are an
important dimension of the edge effect question that needs to enter the discussion.
Moreover, it implies that the apparent relationship between biomass and edge effect
strength may have little to do with the microclimatic explanation given here. To meet the
scholarship standards of CB, such possible alternative interpretations need to be at least
mentioned. The causal explanation presented by the authors is only one of many.
Conclusions on mechanisms are at best speculative based on the relatively weak purely
correlation evidence presented. A coarse generalization that ignores these latitudinal
differences in species guilds and their will not be useful to conservation practitioners, a
usefulness that is an express goal of this paper and CB.
Page 18: This pattern could also be interpreted in alternative ways. For instance, based on
my observations, the low biomass temperate Australian and Florida forests should be
more correctly classified as savannas rather than forest because they do not have a closed
canopy (i.e the whole landscape is edge). The other low biomass forest in the bird dataset
is boreal. As mentioned before, the clear cut edges in the boreal forests are very different
than temperate forest that are fragmenting due to agriculture and development.
Specifically, as the authors mention in the introduction, edge effects on birds are
frequently due to brood parasitism or predation by small predators. The boreal forest
neither has brood parasites like cow birds nor do small and medium sized predators
concentrate along clear cut edges. In contrasst areas subject to higher human population
densities have abnormally high concentrations of small predators (like domestic cats)
which like to hunt along forest edges. These points are important because in the savanna
case the biomass effect results from a complete lack of "forest interior" habitat in the
landscape, and in the other types of forest the mechanism has more to do with predator
and brood parasite behavior than with forest biomass or micro-climate.
Page 19: Note that plants are probably more strongly affected by microclimate than
animals. It is well known that most forest understory herbs do not do well in open
habitats. Plant communities show much more stark contrast between open edge and forest
interior habitat compositions than animal communities. However, even in plants I would
be surprised if the effect of biomass is important. Once you have enough trees to have a
closed canopy (i.e. a forest interior) you will get lots of specialist species regardless of
biomass.
Page 20: This does not mean that they were not impacted by extinctions. Most of the
studies were conducted in areas were large predators are extinct and meso-predators have
greatly increased in abundance as a result. As mentioned above, meso-predators have
important effects on bird and small mammal survival in edge habitats.
Page 21 re: cross biome experiments: I think it would be a waste of money. Biomass has
a spurious correlation at best for the data presented. Most likely there is a threshold
biomass from closed canopy forests to open savannas, but other wise biomass seems to
have little explanatory power. Conservationists already have much better arguments for
conserving large contiguous habitats. For example range requirements for large animals
and minimum viable population size. Knowing the minimum patch sizes
needed in particular environmental contexts will do far more for conservation than
exploring a global relationship between biomass and strength of edge effects. As
someone that works for a conservation NGO, I can’t see any way how such a global
approximation could be useful in practice. The exercise proposed is purely academic. I
find that even academically it has little merit, given that it ignores causal mechanisms.
Page 22: This speculation seems highly doubtful. Dry tropical systems such as savannas
have low AGB, yet have many more large predators and migratory ungulates that require
large conservation areas. The same is true of the tundra and boreal forests. By contrast
the relatively small and sedentary species of tropical rainforest require much smaller
habitat patches, even once allegedly greater edge effects have been taken into account.
Similarly for plants, large scale fire disturbances of low biomass boreal forest require
large tracts of protected land if late successional specialist (i.e. forest interior species) are
to survive in the long term. The implication of the authors that patches can be smaller in
low biomass systems is misleading to conservationists.
Page 22: Tropical ecologists, generally ignore ecological studies from temperate regions,
and in this case would correctly focus only on the tropical forest studies. This point also
calls into question the motivation for the synthesis attempted in this regression analysis.
Why use a coarse global generalization, if you can use local studies as a much more
accurate guideline?
Page 22: Does “cold” mean “temperate”. The authors use of terminology is inconsistent
and therefore confusing. My observation of edges in temperate forest is that these edge
habitat are as if not more degraded as forest edges in tropical rainforests. This
degradation is partly the result of excessive vine growth that can lead to a reverse
succession and partly a result of large amounts of invasive plant species. Forest interior
plants like spring wildflowers are among the first to disappear along edges in temperate
deciduous forests. In fact edges in temperate deciduous forest can become almost entirely
composed of a diverse assemblage of non-native species, an effect that appears to be less
pronounced in the tropics although that may change as tropical nations become more
engaged in globalization. The authors choice to look at one aspect of fragmentation in
isolation results in misleading conclusions. The fact that these conclusions are moderated
by using tentative wording does not make them any more useful.
Page 23: The BAH is a good example of making a broad generalization that is misleading
because it is based on pure correlation and ignores differences in underlying mechanisms.
For instance if the context is an agricultural edge in the tropics, or the edge of a highway
in the temperate zone or a clear cut edge in the boreal forest matters. The BAH does not
get at these important differences in context. Pointing out that edge effects are a
problem in the tropics is not novel. What conservation biologists working in tropical
rainforests need to know is what they can do about it. CB should focus on papers that
provide solutions that can be exported to other parts of the world.
Table 1: Boreal deciduous forest is simply the pioneer stage that eventually becomes
boreal conifer forest. In fact most of these stands are actually mixed forests. I would
eliminate the distinction between boreal deciduous forest and boreal conifer forest. The
best way to do this is to keep only the BDC because this study focuses on edges at
late seral sites
Table 1: Should either divide the table into sections according to environmental variable
measured or write the environmental variables into the column for each study. Current
presentation is awkward and unconventional. I favor the later so that each study is listed
only once, thereby reducing table size.
Table 1: I doubt that all these boreal stands had the same AGB. Where does the value 63
come from? I recommend combining the three data points of the Messier study with a
single average MEI. Right now it seems “pseudo-replicated”.
Table 1: New Zealand forests are temperate rainforests. Northern New Zealand is at best
subtropical. It is inappropriate to lump these forest in with tropical rainforests.
Table 2: The edges in these studies include hard edges like agriculture in Queensland
Australia to forest clear cuts in eastern Canada. The effect on edge effects and species is
very different. A regenerating clear cut provides good habitat with natural forest
vegetation whereas dairy pastures do not. This difference creates a bias because northern
sites tend to involve clear cuts while southern sites tend to involve cattle pastures.
Reviewer # 2:
These authors attempt to address a key question in landscape ecology; what factor(s)
explain the substantial observed variation in edge effects across regions and studies? As
such the topic is certainly one that is of value to readers of Conservation Biology. The
authors summarize literature from three broad taxa: mammals, birds, beetles to test two
predictions flowing from the ‘biomass accumulation hypothesis’ (a hypothesis forwarded
by some of the authors in previous work).
Major concerns:
1.
Given the messy nature of the data used to address the question (see points on
confounding variables below) it is highly surprising that the authors found such a strong
signal. A key factor to be addressed in most meta-analyses is whether the studies selected
represent an unbiased representation of the literature and that there was not significant
publication bias involved. The authors do not present any sort of analysis in this regard.
2.
With such a small number of studies represented it is difficult to generalize to the
broader population of edge studies. Idiosyncratic effects of single studies have
large weight in this analysis. I find it very surprising that the authors were able to find
many studies in the high biomass temperate rainforest that revealed any edge effects.
Most studies completed to date on fragmentation effects in western landscapes
(McGarigal & McComb 1995, Bunell 1999, etc.) reveal that fragmentation effects tend to
have much lower magnitude in such ecosystems. I am also struggling to think of many
temperate mammal species that have been quantitatively deemed ‘interior specialists’.
The authors should discuss why their results seem to contrast so strikingly with many
other studies that did not fit their study criteria but that found no edge effects.
3.
As stated, some readers may find the key hypothesis examined to be trivial. To
paraphrase this hypothesis: if the contrast between types is low, the effect of edge will be
low. Because edge = contrast between two types (Authors own definition: “Edge effects
are the result of the interaction between two adjacent ecosystems that are separated by an
abrupt transition”) Thus potential for triviality; More contrasting edge = more contrasting
edge effects. The authors need to be clearer about why this particular hypothesis is novel
and a worthwhile contribution to existing knowledge.
Other comments
‘Interior forest conditions’ is not well defined. See Villard (1998) suggesting that the
concept of ‘interior’ species is not well supported by data.
“High levels of vegetation structure result from high net primary productivity and
infrequent loss of vegetation to disturbance” and/ or?
“This steeper gradient in resources and conditions may be more finely partitioned by
plants and animals, leading to a greater percentage of species specializing on forest
interior, edge, or disturbance patch interior conditions.” This hypothesis is not clear as
stated in this part of the manuscript. Why does the steepness of a gradient have anything
to do with the degree to which partitioning occurs? Can’t a high degree of partitioning
happen across very gradual gradients? Indeed, more gradual gradients would allow for
more geographic space within which to ‘fit’ species that specialize in these various
conditions (assuming true ‘edge’ species exist; see Villard 1998).
“In contrast, edges in ecosystems with low AGB are predicted to have less of a buffering
effect; microclimatic gradients and other processes will also be less pronounced”
“The hypothesis may provide an framework” change to ‘a’ framework.
P. 14. How could the authors test for a relationship between MEI and AGB with n=3, and
n=4?
“Within this model, AGB was possitively” change to positively
Fig. 2 has no y-axis label.
Fig. 3 regression lines should not extend beyond values for the data.
Fig. 4. With <10 data points in the regression there is probably no need to have r2 and
coefficients with 4 significant digits.
Species response to edge was quantified as the “proportion of species found in forests
that were significantly more abundant in forest interior than at the forest edge”. What
was considered to be the ‘interior’ in this study?
Was it defined individually for each study based on microclimactic data? If so, this
should be more clear.
One key risk to examining one landscape factor in the absence of others is that many of
these variables tend to be confounded in nature. The authors correctly point out that there
are many site-scale confounding factors, but this is only part of the problem. For instance,
Andren (1994 and other studies) showed that edge effect may only matter in landscapes
with low proportions of ‘habitat’. Fahrig and others have pointed out that edge effect,
patch size (configuration variables) tend to be intercorrelated AND correlated with the
amount of habitat at landscape scales. Thus, a key question is whether the effects the
authors are observing are directly a function of edge or perhaps a function of some other
factor correlated with edge (e.g., area effects, habitat loss effects).
Justification for no formal meta-analysis: the authors justify a more ad hoc approach to
summarizing papers based on the fact that statistical reporting in the ‘early phase’ of
conservation biology was not as rigorous. However, most of the studies reported are from
the late ‘90s and early 00’s.
“Consequently, we expect that extinction rates of interior species in small patches will be
higher in high biomass ecosystems.” All other aspects being equal. Edge effects are one
of many mechanisms that drive patch size effects (e.g., stochastic environmental and
demographic processes, minimum home range size effect).
Reviewer # 3:
I offer the following suggestions:
 Shorten significantly to stay focused on the central theme. As it is now, the ms
simply did not hold my attention for its length. I would have liked to read
through it one more time but I just ran out of time.
 Make the process of study selection more transparent.
NOTE: This reviewer made numerous editorial comments and corrections in Track
Changes in Word – it was clear from these edits that considerable more attention needs to
be given to clarity, conciseness, and proof-reading.